Convolutional neural network-based tool condition monitoring in vertical milling operations using acoustic signals

Clayton Cooper; Peng Wang; Jianjing Zhang; Robert X. Gao; Travis Roney; Ihab Ragai; Derek Shaffer

doi:10.1016/j.promfg.2020.07.004

Convolutional neural network-based tool condition monitoring in vertical milling operations using acoustic signals

Clayton Cooper, Peng Wang, Jianjing Zhang, Robert X. Gao, Travis Roney, Ihab Ragai, Derek Shaffer

Research output: Contribution to journal › Conference article › peer-review

40 Scopus citations

Abstract

Sonic monitoring presents itself as one of the least invasive but easiest to implement methods of machine condition characterization. This work investigates the viability of categorically classifying cutting tool wear using only sonic output from a vertical milling center and proposes a statistical model of milling acoustic signals as well as a novel machine learning-integrated method of acoustic signal differentiation. To this end, a deep convolutional neural network is used for data classification. Experimental results support the proposed sonic model and demonstrate that tool wear classification accuracy as high as 99.5% is possible using a two-dimensional deep convolutional neural network.

Original language	English
Pages (from-to)	105-111
Number of pages	7
Journal	Procedia Manufacturing
Volume	49
DOIs	https://doi.org/10.1016/j.promfg.2020.07.004
State	Published - 2020
Event	8th International Conference on Through-Life Engineering Services, TESConf 2019 - Cleveland, United States Duration: Oct 27 2019 → Oct 29 2019

Bibliographical note

Publisher Copyright:
© 2019 The Authors.

Keywords

Acoustic signals
Convolutional neural network
Tool condition monitoring

ASJC Scopus subject areas

Industrial and Manufacturing Engineering
Artificial Intelligence

Access to Document

10.1016/j.promfg.2020.07.004

Cite this

@article{aae067f5cf46444297a78c45709953c7,

title = "Convolutional neural network-based tool condition monitoring in vertical milling operations using acoustic signals",

abstract = "Sonic monitoring presents itself as one of the least invasive but easiest to implement methods of machine condition characterization. This work investigates the viability of categorically classifying cutting tool wear using only sonic output from a vertical milling center and proposes a statistical model of milling acoustic signals as well as a novel machine learning-integrated method of acoustic signal differentiation. To this end, a deep convolutional neural network is used for data classification. Experimental results support the proposed sonic model and demonstrate that tool wear classification accuracy as high as 99.5% is possible using a two-dimensional deep convolutional neural network.",

keywords = "Acoustic signals, Convolutional neural network, Tool condition monitoring",

author = "Clayton Cooper and Peng Wang and Jianjing Zhang and Gao, {Robert X.} and Travis Roney and Ihab Ragai and Derek Shaffer",

note = "Publisher Copyright: {\textcopyright} 2019 The Authors.; 8th International Conference on Through-Life Engineering Services, TESConf 2019 ; Conference date: 27-10-2019 Through 29-10-2019",

year = "2020",

doi = "10.1016/j.promfg.2020.07.004",

language = "English",

volume = "49",

pages = "105--111",

}

TY - JOUR

T1 - Convolutional neural network-based tool condition monitoring in vertical milling operations using acoustic signals

AU - Cooper, Clayton

AU - Wang, Peng

AU - Zhang, Jianjing

AU - Gao, Robert X.

AU - Roney, Travis

AU - Ragai, Ihab

AU - Shaffer, Derek

PY - 2020

Y1 - 2020

N2 - Sonic monitoring presents itself as one of the least invasive but easiest to implement methods of machine condition characterization. This work investigates the viability of categorically classifying cutting tool wear using only sonic output from a vertical milling center and proposes a statistical model of milling acoustic signals as well as a novel machine learning-integrated method of acoustic signal differentiation. To this end, a deep convolutional neural network is used for data classification. Experimental results support the proposed sonic model and demonstrate that tool wear classification accuracy as high as 99.5% is possible using a two-dimensional deep convolutional neural network.

AB - Sonic monitoring presents itself as one of the least invasive but easiest to implement methods of machine condition characterization. This work investigates the viability of categorically classifying cutting tool wear using only sonic output from a vertical milling center and proposes a statistical model of milling acoustic signals as well as a novel machine learning-integrated method of acoustic signal differentiation. To this end, a deep convolutional neural network is used for data classification. Experimental results support the proposed sonic model and demonstrate that tool wear classification accuracy as high as 99.5% is possible using a two-dimensional deep convolutional neural network.

KW - Acoustic signals

KW - Convolutional neural network

KW - Tool condition monitoring

UR - http://www.scopus.com/inward/record.url?scp=85090504006&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85090504006&partnerID=8YFLogxK

U2 - 10.1016/j.promfg.2020.07.004

DO - 10.1016/j.promfg.2020.07.004

M3 - Conference article

AN - SCOPUS:85090504006

VL - 49

SP - 105

EP - 111

JO - Procedia Manufacturing

JF - Procedia Manufacturing

T2 - 8th International Conference on Through-Life Engineering Services, TESConf 2019

Y2 - 27 October 2019 through 29 October 2019

ER -

Convolutional neural network-based tool condition monitoring in vertical milling operations using acoustic signals

Abstract

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this